10:30
Cardiac Diagnostics
Chair: Frans van de Vosse
10:30
15 mins
|
IN VITRO COMPARISON BETWEEN DCE-US AND THERMODILUTION FOR BLOOD VOLUME ASSESSMENT
Giovanna Russo, I.H.F. Herold, H.C. van Assen, H.H.M. Korsten, Massimo Mischi
Abstract: Blood volume (BV) assessment provides valuable information on the circulatory system functionality. Nowadays, BV measurements are performed by employment of indicator dilution techniques, such as thermodilution [1]. Although considered as the gold standard, thermodilution is a very invasive procedure due to the need for central catheterization. Recently, Dynamic Contrast-Enhanced Ultrasound (DCE-US) has been proposed as an alternative minimally-invasive approach for BV assessment [2]. This method measures the BV by using a single peripheral injection of a small bolus of Ultrasound Contrast Agent (UCA) detected by an ultrasound scanner. By measuring the acoustic backscatter, two indicator dilution curves (IDCs) can be derived from two different sites in the circulatory system. IDC analysis permits deriving the mean transit time (MTT) that the injected bolus takes to cover the distance between the two measurements sites. Assessment of the BV between these sites is therefore obtained by multiplying the MTT by the blood flow. The system is validated in-vitro for a wide range of flows and volumes, comparing the proposed method with standard thermodilution.
The in vitro system consisted of a flow generator, an electromagnetic flowmeter to measure and adjust the generated flow, heating devices to keep a constant temperature (37 oC), two thermistors for thermodilution, a tube network simulating the BVs, and a pressure stabilizer.
A small bolus of UCA diluted in cold saline (1-mg SonoVue® in 20 mL saline at 4 °C) was injected into the system. The cold UCA passage through a first and a second region of interest (ROI) was measured simultaneously with an ultrasound transducer and two thermistors (RADI wires). The measurements were performed for different flows and volumes. BVs were estimated by using two different approaches. In the first one, the IDCs were processed and fitted separately with dedicated models to estimate the MTT of the cold saline and the UCA bolus between the two ROIs. In the second one, the BV was quantified by estimation and analysis of the dilution impulse response between the two ROIs.
A linear relation between BVs estimated by the two techniques (thermodilution and DCE-US) was observed with a correlation coefficient of 0.95. The most significant differences between the two techniques were observed in the case of high volume and low flow, possibly due to a different transport kinetics between UCAs and heat. Given the good correlation between BVs estimated with DCE-US and thermodilution, DCE-US seems a valid minimally invasive alternative to thermodilution for BV measurements.
REFERENCES
[1] H. J. C. Swan, et al., N. Engl. J. Med., vol. 283, pp. 251-256 (1970).
[2] M.Mischi, et al, IEEE Trans UFFC, 51(9), 1137-1141 (2004).
|
10:45
15 mins
|
DEVELOPMENT OF A LED-MULTISPECTRAL IMAGING SYSTEM FOR NON-CONTACT TISSUE PERFUSION AND OXYGENATION IMAGING FIRST RESULTS OF CLINICAL INTERVENTION STUDIES
John Klaessens, Herke Jan Noordmans, Martin Nelisse, Ruud Verdaasdonk
Abstract: During clinical interventions objective and quantitative information of the tissue perfusion, oxygenation or temperature can be very useful for the surgical strategy. Spectral point measurements give limited information and affected areas can easily be missed, therefore imaging methods like hyper-spectral imaging and IR thermography1 are required to determine the tissue condition for large areas.
We developed a hyper-spectral imaging system, based on a programmable LED light source (17 different wavelengths between 370 nm and 880 nm). The system was successfully validated under laboratory conditions on volunteers. Different algorithms to study the oxy- and deoxy-hemoglobin changes in tissue over time were developed and applied for different sets of wavelengths2.
This multispectral imaging system and a thermo camera were applied during clinical interventions: Tissue flap transplantations (ENT), local anesthetic blocks and during open brain surgery (to resect the focus of epilepsy). The images covered an area of 20x20 cm.
The measurements in clinical practice turned out to be more complicated compared to the laboratory experiment due to light fluctuations in the environment, movement of patient and limited field of view in operation room. Using self-written MultiSpec and Match software, light fluctuations and movements were corrected. Oxygenation concentration images could be calculated and combined with temperature images.
During brain surgery, changes in oxygenation on the cortex can reveal the effect of an epileptic attack and slow brain waves can be seen running over the cortical surface. The effectivity of local anesthesia of a hand could be predicted in an early stage using the thermal camera and the reperfusion of transplanted skin flap could be imaged.
The LED-based multispectral imaging system combined with thermal imaging provide complementary information on perfusion and oxygenation changes and are promising techniques for real-time diagnostics during clinical interventions.
REFERENCES
1. E.F.Ring and K.Ammer, "Infrared thermal imaging in medicine,", Physiol Meas., vol. 33, pp. R33-R46, (2012/3).
2. J.H.G.M.Klaessens, H.J.Noordmans, R.de Roode, and R.M.Verdaasdonk, "Non-invasive skin oxygenation imaging using a multi-spectral camera system: effectiveness of various concentration algorithms applied on human skin,", SPIE BiOS Biomedical Optics, SPIE Vol: 7174, (2009/2/23).
|
11:00
15 mins
|
IMPAIRED VENOUS RETURN DUE TO RIGHT VENTRICULAR DIASTOLIC DYSFUNCTION AND ATRIAL BACKWARD EJECTION IN PULMONARY ARTERIAL HYPERTENSION
Tim Marcus, Marielle van de Veerdonk, Frank Oosterveer, Anton Vonk Noordegraaf, Ruud Verdaasdonk
Abstract: Introduction
In pulmonary arterial hypertension (PAH), the overloaded right ventricle becomes stiffer, leading to increase of the right ventricular end-diastolic pressure (RVEDP) and the right atrium (RA) pressure. The subsequent impaired venous return may become manifest by reflux in the Vena Cava (VC). The aim of this study was to explore the associations between VC reflux, RA pressure and RVEDP in PAH.
Methods
In 35 patients, the volumetric flow in both the VC superior and inferior was measured using MRI phase-contrast velocity quantification. The fraction of backward flow in the VC was calculated by dividing the total reverse volume by the total forward volume per heartbeat. By invasive right heart catheterisation, the right ventricular end-diastolic pressure (RVEDP) and the right atrial maximum pressure (RAP_max) were recorded. Associations were assessed by linear regression.
Results
For 8 patients, the VC backflow fraction was above 20%, thus severely impairing the effective venous return. For the whole patient group, the VC backflow fraction was related to the RAP_max (p< 0.00001, r=0.73). RAP_max was related to RVEDP (p<0.000001, r=0.81), and the VC backflow fraction was related to RVEDP (p<0.000001, r=0.73).
Discussion
In case of RV diastolic dysfunction, the RA pressure is not able to overcome the increased RVEDP at end-diastole, resulting in partial atrial backward ejection into the low-pressure VC. The backflow volume can become so large, because there are no effective valves between RA and VC. Thereby this backflow phenomenon is hemodynamically very disadvantageous, and may well be the explanation for the prognostic value of RA pressure in PAH, as assessed by Benza et al. (2010).
By MRI, the PAH treatment can be monitored non-invasively, and therapy can be optimized for the individual patient.
Conclusion
Backward flow in the VC is associated with increased right atrial pressure and RV end-diastolic pressure in PAH. This explains the impact of RV diastolic dysfunction on right atrial pressure and subsequently on the venous return in PAH.
Reference
Benza RL et al, Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL). Circulation. 2010 Jul 13;122(2):164-72.
|
11:15
15 mins
|
NONINVASIVE COMPOUND ULTRASOUND ELASTOGRAPHY FOR VULNERABLE PLAQUE DETECTION IN VIVO
Rik Hansen, Gert Jan de Borst, Michiel Bots, Gerard Pasterkamp, Chris de Korte
Abstract: Introduction: Carotid stenosis with plaque rupture is one of the main underlying causes of stroke. Key issue in the risk assessment is identification of the plaque prone to rupture. Lipid-rich and inflammatory plaques with a thin fibrous cap (vulnerable plaques) have a higher risk at rupture than fibrous plaques (stable plaques). We developed a noninvasive ultrasound based compound elastography method to estimate strains in a plaque caused by the pulsating blood [1], because we expect soft lipid-rich plaques to have higher strains than stiff fibrotic plaques. In this in vivo in patient study we investigate the relation between these strain estimates and vulnerable plaque features.
Material and Methods: A Medison Accuvix V10 ultrasound system equipped with an L5-13 linear array transducer was used to acquire ultrasound radiofrequency (RF) data of transverse cross sections of 18 severely stenotic carotid arteries of patients before carotid endarterectomy (CEA; surgical excision of the plaque and inner vessel wall). RF data were acquired at three different insonification angles (0°, +20 and -20 degrees). Radial strains were estimated from the angular RF data using an iterative cross-correlation based search algorithm [2] followed by compounding [1]. The percentage of plaque area with strains above 1.5% was determined. After CEA, segments of the imaged plaque cross section were cut and histologically stained. Based on the staining, the presence of plaque vulnerability features was determined: a thin fibrous cap, a high concentration of lipids and a low concentration of collagen, a high concentration of macrophages, and a lack of smooth muscle cells. The sensitivity and specificity of the strain area parameter for the detection of these features were determined.
Results: The developed compound elastography method revealed a positive relation between strain and each of the investigated histology based parameters of vulnerability. The highest sensitivities and specificities were reached for the detection of lipid content and superficial macrophages: 80% and 70%, and 100% and 100%, respectively.
Conclusions: Within an in vivo setting, we confirmed that noninvasive compound elastography provides strain parameters that increase with plaque vulnerability. At the moment additional patients are being included to validate the findings in a larger population.
REFERENCES
[1] H.H.G. Hansen et al., "Full 2D displacement vector and strain tensor estimation for superficial tissue using beam steered ultrasound imaging," Phys. Med. Biol., Vol. 5, pp. 3201-3218, (2010).
[2] R.G.P. Lopata et al., "Performance evaluation of methods for two-dimensional displacement and strain estimation using ultrasound radio frequency data," Ultrasound Med. Biol., Vol. 35, pp. 796-812, (2009).
|
11:30
15 mins
|
EVALUATION OF 2D AORTA ELASTOGRAPHY TECHNIQUES IN PORCINE AORTAS
Richard Lopata, Marcel Rutten, Frans van de Vosse
Abstract: The size and growth of abdominal aortic aneurysms (AAAs) are currently monitored with ultrasound. Intervention of AAAs is not without risk and is performed only when rupture seems inevitable. However, clinical decision making is based on diameter (growth) measurements, which are retrospective and prone to observer errors. Novel techniques, such as wall stress analysis, dive deeper into the underlying mechanical causes of AAA rupture, but require CT imaging. Ultrasound strain imaging and elastography have been proposed for mechanical characterization of the aortic wall. The benefits are the lack of ionizing radiation and high resolution. Although applied successfully in vivo, the accuracy of these techniques is not easily quantified. In this study, an experimental framework is proposed, closely mimicking in vivo circumstances, to evaluate elastography of the abdominal aorta. The mechanical properties of porcine aorta’s are measured and compared with static experiments.
Three porcine aortas, obtained from a local slaughterhouse, were prepared and kept in phosphate-buffered saline solution (l = 15 ± 2 cm). The vessels were mounted in a hemodynamic setup, in which physiological blood flow and pressure were closely mimicked in both shape and magnitude. The vessels were subjected to 30% pre-stretch in the direction of the vessel axis [1]. Water was circulated with a computer-controlled pump, and blood pressure was regulated by adjusting the peripheral resistance and impedance of the setup. Raw ultrasound data (RF) were obtained in the longitudinal direction of the vessel with a MyLab70 (Esaote Europe, NL), equipped with a 2D linear array transducer (fc = 4-11 MHz, fr = 65 Hz) for several cardiac cycles. The pressure was measured with a pressure sensor (St Jude Medical, USA). Next, the aorta was subjected to static pressures, ranging from 80 to 120mmHg, in 5mmHg increments. RF-data were acquired at each pressure level. The data of three cardiac cycles (dynamic) or three experiments (static) were processed with a 2D coarse-to-fine displacement estimation algorithm [2]. By assuming axisymmetry, the volume change in time was calculated, from which the compliance per length (Cl), distensibility (D) and Young’s modulus (E) were estimated.
The average parameter estimates for all three aortas, as measured during the dynamic experiments, were Cl = 3.0 - 3.8 x 10-9 m2/Pa, D = 124 - 175 kPa-1 and E = 359 - 525 kPa. An average intra-subject variability of 5.7 - 5.9% was found for all parameters. The static measurements were only successful for aortas 1 en 2. However, a good correspondence was found between the dynamic (E1 = 359 kPa; E2 = 453 kPa) and the static measurements (E1 = 366 kPa; E2 = 424 kPa). The porcine aortas appear to be stiffer than those of normal human volunteers, which could be caused by the freezing process, denaturation or small leaks. In future work, these in vitro experiments will be validated further by means of mechanical testing.
|
11:45
15 mins
|
ROBUST MOTION CORRECTION IN THE FREQUENCY DOMAIN OF CARDIAC MR STRESS PERFUSION SEQUENCES
Vikas Gupta, Martijn van de Giessen, Hortense Kirisli, Sharon Kirschbaum, Emile Hendriks, Wiro Niessen, Boudewijn Lelieveldt
Abstract: To detect perfusion abnormalities at an early stage of CAD, myocardial perfusion is often assessed by analysing cardiac MR perfusion (CMRP) images. A combination of rest and stress-induced perfusion allows assessing the ability of the heart to adapt to physical exercise, quantified as the myocardial perfusion reserve index (MPRI). However, especially in stress MR acquisitions, the inability of a patient to breath-hold may lead to misalignments between subsequently acquired frames and MPRI, which is based on dynamic contrast uptake (upslope), cannot be measured reliably. Here, we propose a novel motion correction method which is especially aimed at robustness.
Motion artifacts manifest themselves as sudden intensity changes over time and show up as high frequency content. We propose to minimize this high frequency content directly by translating all the frames in the sequence, thereby removing the motion artifacts. By writing the necessary discrete time Fourier transforms as matrix multiplications, an efficient solution strategy is derived.
A dataset comprising rest and stress images (MRI, 1.5 Tesla) from 10 patients with suspected CAD was used to validate the proposed motion correction method. The registration accuracy of the method was assessed based on annotated myocardium contour locations and clinically relevant parameters (relative upslope, MPRI). The proposed method is also compared to an existing method based on independent component analysis (ICA) [1].
Mean displacements in the non-registered sequences were 2.46 (rest) and 4.85 (stress) pixels (average pixel size: 1.52 mm isotropic). For the proposed method (FT), these decreased to 0.15 and 0.23 pixels, respectively. However, for the ICA based method these were about 1.76 and 5.08 pixels, a motion increase for the stress sequences.
Rest and stress upslope parameters of the proposed method (FT) and the ICA method were compared to expert annotations and showed good agreement between FT and expert (not statistically significantly different, level P<0.05), while ICA and experts tended to agree less (P=0.026). ICA mainly failed on stress sequences with large motions. MPRI values showed good agreement between FT and experts.
With minimal user intervention (ROI selection in 1 frame), sequences of 50 frames can now be registered automatically in 20 seconds compared to approximately 1 minute required by ICA and 10 minutes required for manual annotation. To our knowledge, the minimal user effort combined with the robustness of the proposed method make it feasible for the first time to process stress sequences in a clinical setting and use MPRI in patient care.
REFERENCE:
Milles, J., van der Geest, R.J., et al.: Fully automated motion correction in first-pass myocardial perfusion MR image sequences. IEEE TMI 27(11) (2008) 1611–21
|
|
